Job TitleML Operations EngineerJob Level / GradeP3 or P4DivisionAI & Agent Operations (Day 2), AI OrganisationReports ToML Operations Lead (India)LocationIndiaRole FocusTraditional data-science / ML models — NOT LLM / GenAISECTION A: POSITION SUMMARYState the objective and purpose of the role.

Operate the organisation’s traditional data-science / ML models in production as part of the India ML operations team — executing deployments and retraining, monitoring for drift, maintaining lineage, and keeping predictive models (such as churn, fraud and network forecasting) accurate and safe. A hands-on execution role working to the standards and approvals set by the ML Operations Lead. This role is classic MLOps and does not cover LLM / GenAI agents.

SECTION B: KEY RESPONSIBILITIES AND RESULTSIndicate key responsibilities and performance indicators of this role.

#Key Responsibilities & Expected Results1Execute model deployments and retraining through the pipeline under approved plans — packaging, validation runs, staged promotion.2Run drift monitoring (data / feature / concept) and produce the associated reporting; flag threshold breaches.3Validate models against the incumbent (backtest / champion-challenger) before promotion — no blind swaps.4Run optimization algorithms to select best intervention candidate against model predictions.5Maintain the lineage records (version, data, baseline, last-known-positive) and keep them current.6Execute rollbacks to a previous model / safe baseline on instruction.7Produce model-performance, cost and quota reporting for the portfolio.8Raise data-quality issues to the lead and to Data Engineering promptly, with evidence.SECTION C: MAJOR CHALLENGES / TYPICAL PROBLEMS ENCOUNTEREDList the principal challenges or problems faced by the role in achieving the results of the position.

- Business driven targeted changes on intervention results produced by the models.
- Model retraining and rerun cycles based on business updates.
- Keeping models accurate as data and the world change; catching drift early.
- Dependence on upstream data quality.

SECTION D: DECISION MAKING AUTHORITYThe scope and impact of the job and the extent of authority to manage resources and make decisions.

Decisions made under own authorityDecisions referred to higher authority• Deployment, retraining and monitoring execution within approved plans.

- Lineage and reporting upkeep.
- Timely, evidenced escalation of data issues.
- Model-promotion approval (the ML Operations Lead).
- Retraining strategy and threshold changes.
- Rollback decisions beyond agreed procedures.

SECTION E:



QUALIFICATIONS / EXPERIENCE / KNOWLEDGE REQUIREDKey knowledge and skills required for this role, and a suitable level of qualification where applicable.

CategoryEssential for this roleGood to haveEducation & Qualifications• Degree in Computer Science, Data Science, Engineering, or a related field.• ML / cloud certificationsWork Experience• 4–7 years in MLOps / DevOps / data or ML engineering

- Hands-on experience with churn predictive models and CI/CD pipelines and production model deployment.
- Drift monitoring / model-performance reporting experience
- Telco or large-scale data environment

Technical / Professional Skills• Python and CI/CD tooling

- Model deployment, registry and lineage concepts
- Data processing and reporting
- Understanding of model validation and drift
- Champion-challenger / backtesting
- Cloud ML services

Non-Technical / Soft Skills• Precision, reliability and process discipline

- Clear status communication across a cross-border team
- Collaboration with data teamsOther Task-Specific Knowledge• Understanding of traditional predictive model concepts• Awareness of telco use cases (churn, fraud, forecasting)

Job TitleML Operations EngineerJob Level / GradeP3 or P4DivisionAI & Agent Operations (Day 2), AI OrganisationReports ToML Operations Lead (India)LocationIndiaRole FocusTraditional data-science / ML models — NOT LLM / GenAISECTION A: POSITION SUMMARYState the objective and purpose of the role.

Operate the organisation’s traditional data-science / ML models in production as part of the India ML operations team — executing deployments and retraining, monitoring for drift, maintaining lineage, and keeping predictive models (such as churn, fraud and network forecasting) accurate and safe. A hands-on execution role working to the standards and approvals set by the ML Operations Lead. This role is classic MLOps and does not cover LLM / GenAI agents.

SECTION B: KEY RESPONSIBILITIES AND RESULTSIndicate key responsibilities and performance indicators of this role.

#Key Responsibilities & Expected Results1Execute model deployments and retraining through the pipeline under approved plans — packaging, validation runs, staged promotion.2Run drift monitoring (data / feature / concept)



and produce the associated reporting; flag threshold breaches.3Validate models against the incumbent (backtest / champion-challenger) before promotion — no blind swaps.4Run optimization algorithms to select best intervention candidate against model predictions.5Maintain the lineage records (version, data, baseline, last-known-good) and keep them current.6Execute rollbacks to a previous model / safe baseline on instruction.7Produce model-performance, cost and quota reporting for the portfolio.8Raise data-quality issues to the lead and to Data Engineering promptly, with evidence.SECTION C: MAJOR CHALLENGES / TYPICAL PROBLEMS ENCOUNTEREDList the principal challenges or problems faced by the role in achieving the results of the position.

- Business driven targeted changes on intervention results produced by the models.
- Model retraining and rerun cycles based on business updates.
- Keeping models accurate as data and the world change; catching drift early.
- Dependence on upstream data quality.

SECTION D: DECISION MAKING AUTHORITYThe scope and impact of the job and the extent of authority to manage resources and make decisions.

Decisions made under own authorityDecisions referred to higher authority• Deployment, retraining and monitoring execution within approved plans.

- Lineage and reporting upkeep.
- Timely, evidenced escalation of data issues.
- Model-promotion approval (the ML Operations Lead).
- Retraining strategy and threshold changes.
- Rollback decisions beyond agreed procedures.

SECTION E: QUALIFICATIONS / EXPERIENCE / KNOWLEDGE REQUIREDKey knowledge and skills required for this role, and a suitable level of qualification where applicable.

CategoryEssential for this roleGood to haveEducation & Qualifications• Degree in Computer Science, Data Science, Engineering, or a related field.• ML / cloud certificationsWork Experience• 4–7 years in MLOps / DevOps / data or ML engineering

- Hands-on experience with churn predictive models and CI/CD pipelines and production model deployment.
- Drift monitoring / model-performance reporting experience
- Telco or large-scale data environment

Technical / Professional Skills• Python and CI/CD tooling

- Model deployment, registry and lineage concepts
- Data processing and reporting
- Understanding of model validation and drift
- Champion-challenger / backtesting
- Cloud ML services

Non-Technical / Soft Skills• Precision, reliability and process discipline

- Clear status communication across a cross-border team
- Collaboration with data teamsOther Task-Specific Knowledge• Understanding of traditional predictive model concepts• Awareness of telco use cases (churn, fraud, forecasting)

📌 ML Operations Engineer (Pune)
🏢 NCSi
📍 Pune